The provision of streamed video data from a transmission point to viewing or monitoring computers is dominated by internet broadcasting systems. Examples of this technology include NetShow™ by the Microsoft Corporation and RealVideo™ by RealNetwork. Due to the relatively limited bandwidth presently available to most users of the internet, these broadcast systems rely strongly on the use of high level compression algorithms. Some systems rely on compression standards such as those defined by the various Motion Picture Experts Group (“MPEG”) standards. In other cases, proprietary standards have been developed that better suit the compression requirements of live, or at least streamed, video.
Other fields in which streaming video, generally, has been of interest are those of industrial control and automation, and security and access control. A typical example is a casino, in which intensive surveillance can use over 1,500 cameras monitored by a team of, say, 6 operators. Currently, these highly visually intensive applications utilise analog, full bandwidth video and, as required, audio. Any recording that takes place is implemented on analog or digital video recorders, the tapes or other recording media then being archived for a predetermined length of time to ensure that they are available if required. Unfortunately, archiving of video-taped data in this way is unappealing because it requires an entire dedicated analog recording and replay system for multiple channels of video. Moreover, location and replay of a particular time or event associated with one or more cameras can be inconvenient due to the need to retrieve specific tapes and to fast forward and rewind the tapes manually to locate a particular entry points.
In addition traditional CCTV systems rely on large, expensive matrix switchers connected via expensive video cabling to bring CCTV monitoring to operators. Integration with security systems is reserved for the top end of the market. These systems have been largely inflexible because installing and commissioning a new camera or operator monitor is expensive, time consuming, and can involve significant construction works.
A significant constraint of the traditional CCTV system is the method of recording video to VCR tapes. These devices allow read and write operations, but not at the same time. So if a user wishes to view a recently recorded event they must halt any current recording and rewind the tape before being able to view the incident. The only traditional CCTV solution to this problem was a complex method of switching recording to standby VCRs that resulted in recordings being spread across several tapes.
Typical analogue systems are centred around a matrix switcher which provides camera switching and control for the system as shown in FIG. 1. There are inherent disadvantages with this architecture as follows:
Firstly, star cabling configuration using application specific coaxial cabling. This is inherently inflexible and costly in that if an additional camera or monitor is required, a dedicated cable needs to be installed that connects back to the matrix switcher.
Secondly, significant hardware real estate is required for such matrix switcher equipment.
Thirdly, such architectures provide limited recording capability which is restricted by the length of VHS tape. By way of example, a typical recording configuration would be to connect 16 cameras to a multiplexer, which in turn is connected to a video recorder that records in time lapse mode for 24 hours.
The 16-input multiplexer is capable of real time (25 fps in PAL format) recording overall. With 16 cameras being recorded simultaneously, each camera is recorded at approximately 1.5 fps, when recorded on a VCR tape in non time-lapse mode. Using a four hour VCR tape in 24 hour time lapse mode, this frame rate decreases by a factor of 4—recording is now at a rate of one frame every four seconds.
Fourthly, cameras cannot be individually configured with different recording frame rates.
Fifthly, the only way to improve the frame rate is to reduce the number of cameras or the duration of recording. This of course means either adding further VCRs (with a consequent increase in multiplexers, cost and control room real estate) or increasing the frequency of tape changes (which creates more labour for those responsible and an increase in storage space required.
Sixthly, recording and playback of a tape cannot be done simultaneously. To view a tape that is being recorded, the tape has to be stopped, rewound and played.
Finally, due to the mechanical components of a VCR, maintenance requirements are high.
A feature of streamed internet-based video is that the compression algorithms used to minimise bandwidth introduce a delay. The delay can be fixed or variable, depending upon the implementation, and can extend to many seconds in duration. Moreover, internet streaming video transmission standards usually emphasise coherency of sound at the expense of video coherence, since humans are more susceptible to information loss through interrupted speech than the loss of a few frames from full motion video. For example, both NetShow and RealVideo are tuned to give priority to the audio stream over the video stream. This is ideal for content such as news broadcasts, where maintaining an unbroken audio stream is of greater relative importance to human comprehension than the quality of the associated video stream. However, in the security and process control field, movement can be considerably more important than sound. For example, a criminal may take less than a second to break a window. If several frames have been dropped to cope with relatively low available bandwidth, then the crime can be missed. This makes NetShow, and RealVideo and similar products designed for internet video streaming unsuitable to the security and process control markets.
Another issue is that internet broadcasting of live events does not need to be absolutely live. In particular, the audience will not be aware if an event has been delayed by a few seconds. Both NetShow and RealVideo take advantage of this fact to delay broadcast so that the respective systems can determine whether a video stream to be transmitted can further be compressed. For this reason, both products introduce a delay in the order of several seconds.
By contrast, in a process control environment, video is often used to confirm operator control. For example, an operator in a control room can use visual feedback from a video camera to enable him to remotely control pouring of steel. It will be appreciated that any delay between the event occurring and the corresponding video reaching the operator controlling the event can be dangerous.
Similar principles apply to security. A prison guard remotely opening a door needs to know immediately if an escapee is hiding behind the door. Delays of several seconds in this situation are unacceptable.
It is an object of the invention to overcome or at least substantially ameliorate one or more of the disadvantages of the prior art.
In addition, the present invention has a number of non-limiting advantages, as follows:
Firstly, it provides tight integration with Honeywell's Enterprise Buildings Integrator (EBI) software to provide significant ease-of-use. The ability to view live and recorded video from the same operator stations as used for the security management, including new and future stations is a significant advantage. The ability to use any event in the system as a trigger for video recording provides you with the ultimate in flexibility and the association of video to system alarms makes it relatively simple for the user to investigate incidents.
Secondly, it is flexible. Because the present invention transmits compressed video signals via the network it easy to move existing cameras and install new cameras. New CCTV monitors only require a PC with suitable software and a network connection.
Thirdly, it is cost-effective. By utilising industry standard TCP/IP networking, the present invention can share network structures with other corporate groups thus significantly reducing the ongoing cost of CCTV system support. This also means that the cost of adding or relocating cameras is greatly reduced since access to a network connection is generally far easier than having to install dedicated cabling back to a matrix switcher. Further, the present invention works with standard, open, PC and network components and so allows customers to choose their desired components. And since pan/tilt/zoom (PTZ) control can be performed from EBI, no costly matrix switcher is required.
Fourthly, the system is open. The present invention uses non-proprietary hardware to create an open surveillance system. This openness means that a user's initial investment is protected and that the system can be cost effectively expanded at a later date to easily handle additional cameras, incorporate new technology, or to handle changes in building use or configuration.
Fifthly, it is scalable. Because of it's revolutionary architecture, the present invention is capable of supporting both small and large installations of CCTV cameras.
Sixthly, it provides significant user security. The present invention contains operator security features previously only available in high-end security systems like Honeywell Enterprise Buildings Integrator (EBI). Security managers can therefore ensure that individual users are controlled in what they can see and what they can do.
Seventhly, the system enables intelligent recording. The combination of event activated, user activated and scheduled recording means that the user only needs to record the video they want. This not only optimises the use of their storage resources, it means that they don't need to spend endless hours searching for recorded incidents.
Eighthly, it includes an advanced search capability. The present invention's search capabilities use the latest database technologies to ensure the user can quickly find and view any incident.
Ninthly, the system provides advanced archiving so the user never loses the important data. Data is written to DAT tape or other storage media for long-term storage. Users can specify how long each individual section of video will be retained.
Tenthly, it provides smooth playback. The format of recorded video on traditional CCTV systems does not have a time base. This means that the playback speed of incidents tends to vary based on how busy the recording machine is. The recorded video will show a person walking across a field of view at a constant rate as slowing down and then hurrying. By using time indexing when storing video, the present invention is able to replay incidents as they occurred—not in fast or slow motion.